
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import tensorflow as tf
plt.rcParams['figure.figsize'] = [12, 6]
plt.rcParams.update({'font.size': 12})


df = pd.read_excel('columns.xlsx')
col = df['Symbol'].values
col = col.tolist()


df_stableFeedFlow = pd.read_csv(r'https://raw.githubusercontent.com/AshuPraja13/FCC-abnormality-detection/main/NOC_stableFeedFlow_outputs.csv',header=None)
df_stableFeedFlow.set_index = df_stableFeedFlow.iloc[:,0]
df_stableFeedFlow = df_stableFeedFlow.drop(columns=0)
df_stableFeedFlow.columns= col
df_stableFeedFlow.sample(5)


df_varyingFeedFlow=pd.read_csv(r'https://raw.githubusercontent.com/AshuPraja13/FCC-abnormality-detection/main/NOC_varyingFeedFlow_outputs.csv',header=None)
df_varyingFeedFlow.set_index = df_varyingFeedFlow.iloc[:,0]
df_varyingFeedFlow = df_varyingFeedFlow.drop(columns=0)
df_varyingFeedFlow.columns= col
df_varyingFeedFlow.sample(5)


df_sensorDrift = pd.read_csv(r'https://raw.githubusercontent.com/AshuPraja13/FCC-abnormality-detection/main/Fhn_sensorDrift_outputs.csv',header=None)
df_sensorDrift.set_index = df_sensorDrift.iloc[:,0]
df_sensorDrift = df_sensorDrift.drop(columns=0)
df_sensorDrift.columns= col
df_sensorDrift.sample(5)


df_sensorDrift.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1440 entries, 0 to 1439
Data columns (total 46 columns):
 #   Column       Non-Null Count  Dtype  
---  ------       --------------  -----  
 0   F3           1440 non-null   float64
 1   Tatm         1440 non-null   float64
 2   T1           1440 non-null   float64
 3   P4           1440 non-null   float64
 4   deltaP       1440 non-null   float64
 5   P6           1440 non-null   int64  
 6   Fair         1440 non-null   float64
 7   T3           1440 non-null   float64
 8   T2           1440 non-null   float64
 9   Tr           1440 non-null   float64
 10  Treg         1440 non-null   float64
 11  Lsp          1440 non-null   float64
 12  Tcyc         1440 non-null   float64
 13  Tcyc - Treg  1440 non-null   float64
 14  Cco,sg       1440 non-null   int64  
 15  Co2,sg       1440 non-null   float64
 16  P5           1440 non-null   float64
 17  V4           1440 non-null   float64
 18  V6           1440 non-null   float64
 19  V7           1440 non-null   float64
 20  V3           1440 non-null   float64
 21  V1           1440 non-null   float64
 22  V2           1440 non-null   float64
 23  Frgc         1440 non-null   int64  
 24  Fsc          1440 non-null   int64  
 25  ACAB         1440 non-null   float64
 26  AWGC         1440 non-null   float64
 27  F5           1440 non-null   float64
 28  F7           1440 non-null   float64
 29  Fsg          1440 non-null   float64
 30  FV11         1440 non-null   int64  
 31  P1           1440 non-null   float64
 32  P2           1440 non-null   float64
 33  FLPG         1440 non-null   float64
 34  FLN          1440 non-null   float64
 35  FHN          1440 non-null   float64
 36  FLCO         1440 non-null   float64
 37  FSlurry      1440 non-null   float64
 38  FReflux      1440 non-null   float64
 39  Tfra         1440 non-null   float64
 40  T10          1440 non-null   float64
 41  T20          1440 non-null   float64
 42  V9           1440 non-null   float64
 43  V8           1440 non-null   float64
 44  V10          1440 non-null   float64
 45  V11          1440 non-null   float64
dtypes: float64(41), int64(5)
memory usage: 517.6 KB


df_sensorDrift.describe().T


sns.heatmap(df_sensorDrift.corr(),cmap='coolwarm')

<AxesSubplot:>


for n,i in enumerate(df_stableFeedFlow.columns):
    plt.figure(figsize=(12,2))
    plt.plot(df_stableFeedFlow[i])
    plt.xlabel('time (mins)')
    plt.ylabel(i)
    plt.title(df[df['Symbol']==i]['Description'].values)
    plt.show()


for n,i in enumerate(df_varyingFeedFlow.columns):
    plt.figure(figsize=(12,2))
    plt.plot(df_varyingFeedFlow[i])
    plt.xlabel('time (mins)')
    plt.ylabel(i)
    plt.title(df[df['Symbol']==i]['Description'].values)
    plt.show()


for n,i in enumerate(df_sensorDrift.columns):
    plt.figure(figsize=(12,2))
    plt.plot(df_sensorDrift[i])
    plt.xlabel('time (mins)')
    plt.ylabel(i)
    plt.title(df[df['Symbol']==i]['Description'].values)
    plt.show()


from sklearn.preprocessing import StandardScaler
ss = StandardScaler()


X = ss.fit_transform(df_stableFeedFlow)


from sklearn.decomposition import PCA
pca = PCA()


X_pca = pca.fit_transform(X)


plt.figure(figsize=(15,6))
sns.set_style('whitegrid')
sns.lineplot(x=list(range(1,47)), y=np.cumsum(pca.explained_variance_ratio_), drawstyle='steps-pre')
sns.lineplot(x=list(range(0,46)),y=np.cumsum(pca.explained_variance_ratio_))
plt.xlabel('Eigen Values')
plt.ylabel('Ratio of Variance')
plt.title('Variance by each Eigen Value')
plt.show()


pca = PCA(n_components=10)
X_pca = pca.fit_transform(X)


plt.figure(figsize=(15,6))
sns.set_style('whitegrid')
sns.lineplot(x=list(range(1,11)), y=np.cumsum(pca.explained_variance_ratio_), drawstyle='steps-pre')
sns.lineplot(x=list(range(0,10)),y=np.cumsum(pca.explained_variance_ratio_))
plt.xlabel('Eigen Values')
plt.ylabel('Ratio of Variance')
plt.title('Variance by each Eigen Value')
plt.show()


X_train = X.reshape(2880,46,1)


# del model


model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(256,return_sequences=True),input_shape=(46,1)))
model.add(tf.keras.layers.Dropout(0.2))
model.add(tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(128,return_sequences=True)))
model.add(tf.keras.layers.Dropout(0.2))
model.add(tf.keras.layers.Dense(1))
model.compile(optimizer='adam', loss='mse', metrics=['mae'])
model.summary()

Model: "sequential"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 bidirectional (Bidirection  (None, 46, 512)           528384    
 al)                                                             
                                                                 
 dropout (Dropout)           (None, 46, 512)           0         
                                                                 
 bidirectional_1 (Bidirecti  (None, 46, 256)           656384    
 onal)                                                           
                                                                 
 dropout_1 (Dropout)         (None, 46, 256)           0         
                                                                 
 dense (Dense)               (None, 46, 1)             257       
                                                                 
=================================================================
Total params: 1185025 (4.52 MB)
Trainable params: 1185025 (4.52 MB)
Non-trainable params: 0 (0.00 Byte)
_________________________________________________________________


model.fit(X_train,X_train,epochs=30)

Epoch 1/30
90/90 [==============================] - 27s 252ms/step - loss: 0.1400 - mae: 0.2014
Epoch 2/30
90/90 [==============================] - 23s 261ms/step - loss: 0.0028 - mae: 0.0362
Epoch 3/30
90/90 [==============================] - 24s 265ms/step - loss: 0.0023 - mae: 0.0328
Epoch 4/30
90/90 [==============================] - 24s 263ms/step - loss: 0.0021 - mae: 0.0311
Epoch 5/30
90/90 [==============================] - 23s 254ms/step - loss: 0.0020 - mae: 0.0295
Epoch 6/30
90/90 [==============================] - 23s 256ms/step - loss: 0.0018 - mae: 0.0285
Epoch 7/30
90/90 [==============================] - 24s 271ms/step - loss: 0.0018 - mae: 0.0278
Epoch 8/30
90/90 [==============================] - 24s 267ms/step - loss: 0.0017 - mae: 0.0273
Epoch 9/30
90/90 [==============================] - 25s 276ms/step - loss: 0.0017 - mae: 0.0270
Epoch 10/30
90/90 [==============================] - 23s 260ms/step - loss: 0.0016 - mae: 0.0264
Epoch 11/30
90/90 [==============================] - 25s 282ms/step - loss: 0.0016 - mae: 0.0262
Epoch 12/30
90/90 [==============================] - 24s 269ms/step - loss: 0.0016 - mae: 0.0259
Epoch 13/30
90/90 [==============================] - 23s 257ms/step - loss: 0.0015 - mae: 0.0255
Epoch 14/30
90/90 [==============================] - 23s 259ms/step - loss: 0.0015 - mae: 0.0253
Epoch 15/30
90/90 [==============================] - 23s 260ms/step - loss: 0.0015 - mae: 0.0251
Epoch 16/30
90/90 [==============================] - 23s 254ms/step - loss: 0.0015 - mae: 0.0253
Epoch 17/30
90/90 [==============================] - 23s 254ms/step - loss: 0.0015 - mae: 0.0249
Epoch 18/30
90/90 [==============================] - 23s 255ms/step - loss: 0.0015 - mae: 0.0249
Epoch 19/30
90/90 [==============================] - 23s 258ms/step - loss: 0.0015 - mae: 0.0248
Epoch 20/30
90/90 [==============================] - 23s 260ms/step - loss: 0.0015 - mae: 0.0247
Epoch 21/30
90/90 [==============================] - 23s 253ms/step - loss: 0.0014 - mae: 0.0245
Epoch 22/30
90/90 [==============================] - 22s 249ms/step - loss: 0.0014 - mae: 0.0243
Epoch 23/30
90/90 [==============================] - 24s 262ms/step - loss: 0.0014 - mae: 0.0246
Epoch 24/30
90/90 [==============================] - 23s 255ms/step - loss: 0.0014 - mae: 0.0244
Epoch 25/30
90/90 [==============================] - 23s 255ms/step - loss: 0.0014 - mae: 0.0242
Epoch 26/30
90/90 [==============================] - 23s 258ms/step - loss: 0.0014 - mae: 0.0243
Epoch 27/30
90/90 [==============================] - 23s 259ms/step - loss: 0.0014 - mae: 0.0245
Epoch 28/30
90/90 [==============================] - 23s 260ms/step - loss: 0.0014 - mae: 0.0243
Epoch 29/30
90/90 [==============================] - 23s 257ms/step - loss: 0.0014 - mae: 0.0241
Epoch 30/30
90/90 [==============================] - 23s 259ms/step - loss: 0.0014 - mae: 0.0243

<keras.src.callbacks.History at 0x2db734dbfa0>


error_ae = []
for i in range(X.shape[0]):
    y_pred = model.predict(X[i].reshape(1,46,1),verbose=None)[0,:,0]
    error_ae.append(np.abs(X[i]-y_pred).sum())
AE_CL = np.percentile(error_ae,99)


pd.Series(error_ae).plot()
plt.hlines(AE_CL,0,len(error_ae),colors='red',linestyles='--')

<matplotlib.collections.LineCollection at 0x2db734a2c10>


X_reconstructed = np.dot(X_pca,pca.components_)
error_pca = X-X_reconstructed
Q_train = np.sum(np.abs(error_pca),axis=1)
Q_CL = np.percentile(Q_train,99)
# Q_train plot with CL
plt.figure()
plt.plot(Q_train, color='black')
plt.plot([1,len(Q_train)],[Q_CL,Q_CL], linestyle='--',color='red', linewidth=2)
plt.xlabel('Sample #')
plt.ylabel('Q metric: training data')
plt.title(f'Q metrix is max: {Q_train.max()} at:{Q_train.argmax()}mins')
plt.show()


lambda_ = np.diag(pca.explained_variance_)
lambda_inv = np.linalg.inv(lambda_)
T_train = np.zeros(X_pca.shape[0])
for i in range(X_pca.shape[0]):
    T_train[i] = np.dot(np.dot(X_pca[i],lambda_inv),X_pca[i].T)
T_CL = np.percentile(T_train,99)
# T2_train plot with CL
plt.figure()
plt.plot(T_train, color='black')
plt.plot([1,len(T_train)],[T_CL,T_CL], linestyle='--',color='red', linewidth=2)
plt.xlabel('Sample #')
plt.ylabel('T$^2$ metric: training data')
plt.title(f'T$^2$ metrix is max: {np.array(T_train).max()} at:{np.array(T_train).argmax()}mins')
plt.show()


cosine = []
ed = []
for i in range(X.shape[0]):
    v1 = X[i]
    v2 = np.dot(X_pca,pca.components_)[i]
    cosine.append(np.dot(v1,v2)/(np.linalg.norm(v1)*np.linalg.norm(v2)))
    ed.append(np.linalg.norm(v1 - v2))
C_CL = np.min(cosine)
E_CL = np.percentile(ed,99)
# pd.Series(ed).plot(color='black')
# plt.plot([1,len(ed)],[E_CL,E_CL], linestyle='--',color='red', linewidth=2)
# plt.show()
pd.Series(cosine).plot(color='black')
plt.plot([1,len(cosine)],[C_CL,C_CL], linestyle='--',color='red', linewidth=2)
plt.xlabel('Sample #')
plt.ylabel('Cosine similarity metric: training data')
plt.title(f'Cosine Similarity')
plt.show()


Q_CL,T_CL,C_CL,E_CL,AE_CL

(4.1231132155190675,
 20.424350352568798,
 0.9427550112367162,
 0.9281694746755795,
 0.34530417881903325)


def Q_test(X,X_pca,pca_components_,Q_CL):
    X_reconstructed = np.dot(X_pca,pca_components_)
    error_pca = X-X_reconstructed
    Q_train = np.sum(np.abs(error_pca),axis=1)
    # Q_train plot with CL
    plt.figure()
    plt.plot(Q_train, color='black')
    plt.plot([1,len(Q_train)],[Q_CL,Q_CL], linestyle='--',color='red', linewidth=2)
    plt.xlabel('Sample #')
    plt.ylabel('Q metric: training data')
    plt.title(f'Q metrix is max: {Q_train.max()} at:{Q_train.argmax()}mins')
    plt.show()
    return error_pca


def T_test(X_pca,explained_variance_,TCL):
    lambda_ = np.diag(pca.explained_variance_)
    lambda_inv = np.linalg.inv(lambda_)
    T_train = np.zeros(X_pca.shape[0])
    for i in range(X_pca.shape[0]):
        T_train[i] = np.dot(np.dot(X_pca[i],lambda_inv),X_pca[i].T)
    # T2_train plot with CL
    plt.figure()
    plt.plot(T_train, color='black')
    plt.plot([1,len(T_train)],[T_CL,T_CL], linestyle='--',color='red', linewidth=2)
    plt.xlabel('Sample #')
    plt.ylabel('T$^2$ metric: training data')
    plt.title(f'T$^2$ metrix is max: {np.array(T_train).max()} at:{np.array(T_train).argmax()}mins')
    plt.show()


def cosine(X,X_transformed,pca_components_,C_CL,E_CL):
    cosine = []
    ed = []
    for i in range(X.shape[0]):
        v1 = X[i]
        v2 = np.dot(X_transformed,pca_components_)[i]
        cosine.append(np.dot(v1,v2)/(np.linalg.norm(v1)*np.linalg.norm(v2)))
        ed.append(np.linalg.norm(v1 - v2))
#     pd.Series(ed).plot(color='black')
#     plt.plot([1,len(ed)],[E_CL,E_CL], linestyle='--',color='red', linewidth=2)
#     plt.xlabel('Sample #')
#     plt.ylabel('Eucledian Distance metric: training data')
#     plt.show()
    pd.Series(cosine).plot(color='black')
    plt.plot([1,len(cosine)],[C_CL,C_CL], linestyle='--',color='red', linewidth=2)
    plt.xlabel('Sample #')
    plt.ylabel('Cosine similarity metric: training data')
    plt.title(f'Cosine Similarity')
    plt.show()


def autoencoder(df_test,CL):
    X_test = ss.transform(df_test)
    error_ae = []
    error_sum = []
    for i in range(X_test.shape[0]):
        y_pred = model.predict(X_test[i].reshape(1,46,1),verbose=None)[0,:,0]
        error_ae.append(np.abs(X_test[i]-y_pred))
        error_sum.append(np.abs(X_test[i]-y_pred).sum())
    error_ae=np.array(error_ae)
    pd.Series(error_sum).plot(color = 'black')
    plt.hlines(CL,0,len(error_ae),colors='red',linestyles='--')
    plt.xlabel('Sample #')
    plt.ylabel('Reconstruction error by Autoencoder')
    return error_ae


X = ss.transform(df_varyingFeedFlow)
X_test = pca.transform(X)


error_pca = Q_test(X,X_test,pca.components_,Q_CL)


T_test(X_test,pca.explained_variance_,T_CL)


cosine(X,X_test,pca.components_,C_CL,E_CL)


error_ae = autoencoder(df_varyingFeedFlow,AE_CL)


X = ss.transform(df_sensorDrift)
X_test = pca.transform(X)


error_pca = Q_test(X,X_test,pca.components_,Q_CL)


T_test(X_test,pca.explained_variance_,T_CL)


cosine(X,X_test,pca.components_,C_CL,E_CL)


error_ae = autoencoder(df_sensorDrift,AE_CL)


#%% Q contribution
error = np.abs(error_pca).sum(axis=1)
cum = []
for index,value in enumerate(error):
    if (value>Q_CL) and (len(cum)<15):
        previous_val = value
        cum.append(value)
        if len(cum) == 15:
            sample = index
            break
    else:
        cum=[]
# sample = ((pd.Series(error_pca.sum(axis=1))-pd.Series(error_pca.sum(axis=1)).shift()).abs()).argmax()
print('Time-',sample,'mins')
error_test_sample = error_pca[sample]
Q_contri = np.abs(error_test_sample) # *error_test_sample # vector of contributions

Time- 434 mins


plt.figure(figsize=[15,4])
plt.bar(['variable ' + str((i+1)) for i in range(len(Q_contri))], Q_contri)
plt.xticks(rotation = 80)
plt.ylabel('Q contributions')
plt.show()


plt.figure(figsize=(15,40))
print('Time-',sample,'mins')
for i,n in enumerate(np.argsort(Q_contri)[:-11:-1]):
    plt.subplot(5,2,i+1)
    plt.plot(df_sensorDrift.iloc[:,n],'blue', linewidth=1)
    plt.xlabel('time (mins)')
    plt.ylabel(df['Symbol'][n])
    plt.title(df['Description'][n])
    plt.show

Time- 434 mins


#%% Autoencoder Error
error = np.abs(error_ae).sum(axis=1)
cum = []
for index,value in enumerate(error):
    if (value>AE_CL) and (len(cum)<15):
        previous_val = value
        cum.append(value)
        if len(cum) == 15:
            sample = index
            break
    else:
        cum=[]
# sample = ((pd.Series(error_ae.sum(axis=1))-pd.Series(error_ae.sum(axis=1)).shift()).abs()).argmax()
print('Time-',sample,'mins')
error_test_sample = error_pca[sample]
Q_contri = np.abs(error_test_sample) # *error_test_sample # vector of contributions

Time- 743 mins


plt.figure(figsize=(15,45))
print('Time-',sample,'mins')
for i,n in enumerate(np.argsort(error_ae[sample])[:-11:-1]):
    plt.subplot(5,2,i+1)
    plt.plot(df_sensorDrift.iloc[:,n],'blue', linewidth=1)
    plt.xlabel('time (mins)')
    plt.ylabel(df['Symbol'][n])
    plt.title(df['Description'][n])
    plt.show

Time- 743 mins

	F3	Tatm	T1	P4	deltaP	P6	Fair	T3	T2	Tr	...	FLCO	FSlurry	FReflux	Tfra	T10	T20	V9	V8	V10	V11
1480	164.80	75.521	461.23	34.4	-6.4	28	2.6774	1560.8	616.00	968.99	...	1643.4	218.64	2940.9	310.98	509.68	628.04	50.669	49.404	49.830	47.218
1244	165.10	77.351	459.94	34.4	-6.4	28	2.6838	1570.5	616.00	969.03	...	1647.1	211.21	2996.0	313.10	510.05	628.41	48.149	49.958	49.234	47.400
1133	164.84	78.371	460.85	34.4	-6.4	28	2.6804	1563.4	616.01	969.04	...	1642.4	214.63	2978.1	313.47	509.78	628.09	47.583	49.593	49.261	47.153
733	165.28	80.011	460.53	34.4	-6.4	28	2.6858	1568.1	616.01	969.02	...	1649.7	206.82	3024.8	315.23	510.06	628.32	45.464	50.073	49.117	47.545
1296	164.85	76.831	460.67	34.4	-6.4	28	2.6819	1564.6	616.00	969.06	...	1645.8	214.10	2986.2	312.63	510.01	628.38	48.742	49.875	49.056	47.327

	F3	Tatm	T1	P4	deltaP	P6	Fair	T3	T2	Tr	...	FLCO	FSlurry	FReflux	Tfra	T10	T20	V9	V8	V10	V11
8754	166.17	76.453	461.41	34.4	-6.4	28	2.7019	1568.2	616.00	969.04	...	1666.9	205.34	3143.5	315.04	511.34	630.11	46.352	52.081	49.984	48.415
9541	164.76	79.656	460.69	34.4	-6.4	28	2.6770	1563.9	616.00	968.99	...	1645.0	220.22	3000.6	314.76	509.79	628.26	46.093	49.747	50.390	47.296
8351	167.63	78.186	460.40	34.4	-6.4	28	2.7234	1583.5	616.00	969.00	...	1689.6	208.17	3402.6	320.51	513.07	632.49	40.988	55.387	52.210	49.563
1663	167.38	77.671	460.72	34.4	-6.4	28	2.7183	1579.9	616.00	968.99	...	1674.3	207.23	3334.8	319.06	512.70	631.96	42.357	54.535	50.895	48.706
9288	165.10	79.945	461.20	34.4	-6.4	28	2.6815	1562.8	616.02	968.97	...	1650.4	229.87	3098.0	316.78	510.49	629.47	44.194	51.018	51.753	47.537

	F3	Tatm	T1	P4	deltaP	P6	Fair	T3	T2	Tr	...	FLCO	FSlurry	FReflux	Tfra	T10	T20	V9	V8	V10	V11
1326	164.93	76.445	460.96	34.4	-6.4	28	2.6806	1563.3	616.00	969.02	...	1643.8	216.84	2967.1	312.09	509.83	628.26	49.408	49.647	49.574	47.229
214	165.06	77.575	460.70	34.4	-6.4	28	2.6839	1565.7	616.00	969.04	...	1647.0	208.72	2977.8	312.84	509.89	628.09	48.269	49.702	48.879	47.412
644	164.66	79.991	460.84	34.4	-6.4	28	2.6751	1562.3	615.99	968.99	...	1644.1	218.94	2983.7	314.63	509.62	627.97	46.108	49.490	50.249	47.263
239	165.18	77.788	460.64	34.4	-6.4	28	2.6829	1566.7	616.00	969.00	...	1648.2	209.58	2987.5	313.18	509.93	628.21	47.917	49.804	49.294	47.475
302	165.05	78.379	460.70	34.4	-6.4	28	2.6806	1565.6	616.00	968.98	...	1648.1	215.07	3006.0	313.93	509.99	628.45	47.075	49.982	50.068	47.460

	count	mean	std	min	25%	50%	75%	max
F3	1440.0	165.012694	1.582447e-01	164.530000	164.890000	165.020000	165.140000	165.480000
Tatm	1440.0	78.342193	1.491834e+00	74.981000	77.191500	78.746000	79.687250	80.070000
T1	1440.0	460.899750	3.939347e-01	459.580000	460.650000	460.910000	461.140000	462.050000
P4	1440.0	34.400000	8.884870e-13	34.400000	34.400000	34.400000	34.400000	34.400000
deltaP	1440.0	-6.400000	1.732550e-13	-6.400000	-6.400000	-6.400000	-6.400000	-6.400000
P6	1440.0	28.000000	0.000000e+00	28.000000	28.000000	28.000000	28.000000	28.000000
Fair	1440.0	2.680681	2.799086e-03	2.672600	2.678675	2.680700	2.682800	2.687600
T3	1440.0	1564.149444	2.865294e+00	1555.800000	1562.200000	1564.100000	1566.000000	1573.500000
T2	1440.0	616.000063	7.019004e-03	615.980000	616.000000	616.000000	616.000000	616.020000
Tr	1440.0	969.000271	2.893517e-02	968.900000	968.980000	969.000000	969.020000	969.090000
Treg	1440.0	1249.999306	4.729886e-02	1249.900000	1250.000000	1250.000000	1250.000000	1250.100000
Lsp	1440.0	29.696215	9.494108e-02	29.463000	29.627000	29.697500	29.767000	29.942000
Tcyc	1440.0	1255.282708	4.426402e-02	1255.200000	1255.300000	1255.300000	1255.300000	1255.400000
Tcyc - Treg	1440.0	5.283424	3.261765e-02	5.197400	5.258700	5.283350	5.308050	5.353900
Cco,sg	1440.0	29878.145139	4.297248e+01	29772.000000	29848.000000	29877.000000	29910.000000	29995.000000
Co2,sg	1440.0	0.012496	1.463217e-04	0.012105	0.012383	0.012495	0.012605	0.012816
P5	1440.0	24.900000	6.752501e-13	24.900000	24.900000	24.900000	24.900000	24.900000
V4	1440.0	46.263644	3.639285e-01	45.379000	46.008750	46.375500	46.558250	46.840000
V6	1440.0	24.800864	8.697047e-02	24.530000	24.747000	24.809000	24.862000	25.002000
V7	1440.0	54.594028	5.706766e-02	54.429000	54.553000	54.595000	54.638000	54.735000
V3	1440.0	46.989394	1.545578e-02	46.949000	46.978000	46.990000	47.001000	47.026000
V1	1440.0	57.939251	2.065033e-01	57.343000	57.801000	57.934000	58.073000	58.619000
V2	1440.0	45.296442	5.164279e-02	45.163000	45.259000	45.294000	45.335000	45.436000
Frgc	1440.0	49586.847222	5.545769e+01	49425.000000	49546.000000	49589.000000	49631.000000	49735.000000
Fsc	1440.0	49586.981944	5.556882e+01	49425.000000	49547.000000	49588.000000	49631.000000	49729.000000
ACAB	1440.0	280.759167	1.341559e+00	277.200000	279.750000	281.150000	281.840000	282.770000
AWGC	1440.0	204.625167	2.503420e+00	198.600000	202.837500	205.400000	206.680000	208.490000
F5	1440.0	1990.668125	7.094397e+00	1970.200000	1985.900000	1990.500000	1995.300000	2014.000000
F7	1440.0	3737.161944	4.866664e+00	3723.800000	3733.700000	3737.100000	3740.800000	3747.800000
Fsg	1440.0	160.840868	1.682502e-01	160.350000	160.720000	160.840000	160.970000	161.260000
FV11	1440.0	28116.102778	2.741912e+02	27445.000000	27925.750000	28201.000000	28338.000000	28547.000000
P1	1440.0	14.638000	7.640988e-14	14.638000	14.638000	14.638000	14.638000	14.638000
P2	1440.0	35.041949	2.131480e-02	35.000000	35.028000	35.038000	35.055000	35.104000
FLPG	1440.0	3037.770347	4.418845e+01	2931.300000	3006.075000	3051.500000	3074.000000	3106.500000
FLN	1440.0	3917.952569	4.534407e+01	3853.300000	3878.675000	3904.800000	3953.225000	4029.600000
FHN	1440.0	724.872389	1.138577e+01	702.770000	715.775000	725.260000	733.930000	751.220000
FLCO	1440.0	1643.657778	3.783916e+00	1632.900000	1640.800000	1643.700000	1646.500000	1652.200000
FSlurry	1440.0	213.914854	3.732455e+00	205.910000	211.110000	213.995000	216.580000	224.070000
FReflux	1440.0	2984.431944	1.881372e+01	2938.600000	2972.300000	2986.900000	2996.825000	3029.200000
Tfra	1440.0	313.547104	1.204784e+00	310.710000	312.657500	313.910000	314.560000	315.330000
T10	1440.0	509.836000	1.324484e-01	509.500000	509.740000	509.840000	509.932500	510.150000
T20	1440.0	628.161542	1.593855e-01	627.840000	628.040000	628.140000	628.282500	628.470000
V9	1440.0	47.511359	1.506424e+00	45.384000	46.213750	47.062000	48.649750	51.247000
V8	1440.0	49.688406	1.921797e-01	49.287000	49.534000	49.699000	49.845000	50.130000
V10	1440.0	49.368132	4.821624e-01	47.873000	49.071750	49.349500	49.692750	50.506000
V11	1440.0	47.220358	2.076930e-01	46.630000	47.065500	47.224000	47.379000	47.687000

Increased Heavy Nhaptha Flow Rate from Fractionator Column¶

Introduction¶

Data Preprocessing¶

Exploratory Data Analysis (EDA)¶

Anomaly Detection Techniques¶

Data Cleaning, Feature Engineering, Predictive Modeling¶

Testing the model on Varying feed flow rate.¶

Testing the model on abnormal dataset.¶

Visualization¶

Conclusion¶

Future Work¶

*Increased Heavy Nhaptha Flow Rate from Fractionator Column*¶

Introduction¶

Data Preprocessing¶

Exploratory Data Analysis (EDA)¶

Anomaly Detection Techniques¶

Data Cleaning, Feature Engineering, Predictive Modeling¶

Testing the model on Varying feed flow rate.¶

Testing the model on abnormal dataset.¶

Visualization¶

Conclusion¶

Future Work¶

Increased Heavy Nhaptha Flow Rate from Fractionator Column¶